This invention generally relates to lasers and is particularly concerned with a novel method and apparatus by which the stability of the amplitude and repetition rate of mode-locked laser pulses can be controlled.
The stability of the amplitude and repetition rate of mode-locked laser pulses, such as Nd:YAG laser pulses, is of importance in pulsed laser communication systems. When a pulsed laser communication system operates in a background limited condition, the signal-to-noise ratio at the receiver can be maximized by gating the receiver and by setting a detection threshold level at the receiver, as is known. The accuracy with which the receiver can be gated or a threshold level set depends on how well the mode-locked pulses can be stabilized.
When a laser is mode-locked, its cavity length has to be matched to the frequency of the output pulses and this condition is given approximately by EQU w.sub.m = c/ 2L, (1)
where w.sub.m is the frequency of the mode-locked pulses, c is the speed of light, and L is the optical path length of the laser cavity. Thermal effects in the laser components are usually the primary cause of laser cavity length variations, and thus the problem of stabilization becomes one of continuously compensating for these thermal perturbations.
Two methods of stabilizing mode-locked lasers are currently known. The first method utilizes a simple feedback loop which detects the mode-locked pulses and then uses the amplified detector output to drive the laser mode-locker. The second known method utilizes a phase-lock loop which compares the detected output of the laser to the signal applied to the mode-locker, the resultant error signal being used to correct the drive of the mode-locker. Both of these known methods have similar disadvantages, however, as neither can easily be used with a loss modulator mode-locking element since time varying losses inside the cavity introduce relaxation oscillation noise on the laser output which, in turn, produces noise in the feedback loops. When a loss modulator is driven hard enough to produce short mode-locked pulses, this noise can increase to a level which causes these feedback systems to lose lock. Both known methods also allow the transmitter frequency to vary continuously, which limits the accuracy with which the detector can be gated.